JAMES BUCHANAN is without doubt one of the most influential economists of our time. As the main founder of the field of public choice or the new political economy he has created a new approach to the study of public economics and economic policy. The effects of his efforts and those of his many collaborators--among whom of course Gordon Tullock is the most outstanding--are striking. Public choice has come to be recognized as a distinct approach to the study of economic policy and has been developed by a large number of researchers both in the United States and in the rest of the world. A specialized journal and an international society have been founded to stimulate work in the area, but contributions to public choice have also come to permeate general economics journals and conferences. Moreover, even those who do not consider themselves "public choice economists" have in significant ways been influenced by the approach. In particular, Buchanan's views on the relationship between economics and politics have had an important and probably enduring influence on the profession in general.
As in the case of many other important contributors to the world of ideas Buchanan's name is often associated with theories, ideas, and attitudes in a rather loose way. Even those economists who have never read anything of what he has written (assuming that such persons exist) may feel that they have a fairly good idea of what his work is all about. The reason for this is of course that his name has become synonymous with the school of thought that he has founded, and as the ideas of the school have spread, the many references to Buchanan's books and articles may have left a rather vague impression that "it is all in Buchanan."
This is certainly not true. If one reads the modern literature on public choice it is easy to find contributions that in basic ideas in research methodology, and in the political implications drawn are quite different from what one finds in the founder's work, but one can also find in his work ideas that have not yet been taken up by his followers. This makes it all the more interesting to find out what actually are the ideas that can be ascribed directly to Buchanan. This is no easy task. The bibliography supplied by Buchanan himself to the issue of the Scandinavian Journal of Economics marking the occasion of the award of the 1986 Nobel Memorial Prize in Economics, contained 374 items, including 37 books and monographs, and the number must have increased substantially since then. To read it all obviously requires an unusually keen interest both in the subject and in the man. Fortunately, the two books under review here(n1) provide us with an excellent selection of Buchanan's articles which, as far as I can judge, also gives a representative view of his contributions to economics, including the economics of politics and the politics of economics.
The two books differ somewhat in scope and purpose. Economics: Between Predictive Science and Moral Philosophy (hereinafter referred to as Economics) is a broad selection of Buchanan's articles from the early 1950s to the mid-1980s. The editors of this volume, Robert D. Tollison and Viktor J. Vanberg, say that "we believe that our selection does not depart too much from what the author himself considers to be his major essays." Liberty, Market and State (hereinafter referred to as Liberty) is Buchanan's personal selection of articles published in the early 1980s and reflects his main preoccupations during that period. The focus is therefore somewhat narrower--although "narrow" must be understood in a relative sense, because the issues considered are certainly very broad and of general interest. Two articles are included in both volumes, yielding a net number of 48 articles in two books together.(n2)
The books are nicely produced and edited, although I have one criticism to make of the editors of Economics. Because the readers of such a book must be presumed to take some interest in the development of Professor Buchanan's thinking and its relationship to the main intellectual currents of the times, they ought to be told the year of the first publication of each individual article. In about half the cases the editors' acknowledgment of the permission to reprint refers only to the volume of the journal in question, and the reader will have to find the rest for himself. In this respect Professor Buchanan himself is a model editor.
Many reviewers of collections of articles choose the strategy of concentrating on a few articles that they find especially interesting and then commenting on these in some detail. This would not be a natural strategy to follow in the case of Professor Buchanan. He is not the type of economist who wanders from one field to another and applies the methodology that he masters to the various interesting problems that he picks up as he goes along. There is a unity of interest and purpose in his writings that makes it natural to comment on his work more by theme than by article, and this is the way in which I have chosen to organize this review. In the next section I present a brief summary of the contents of the two books. I then pick out some particularly important ideas in his work and comment on them in more detail in Sections III-V. The final section includes some further reflections on Buchanan's work.(n3)
II. Main Themes
Each of the two books is divided into five parts with four to six articles in each. Part 1 of Economics is entitled "What Should Economists Do?" The 1959 article, "Positive Economics, Welfare Economics, and Political Economy," gives a very clear early statement both of Buchanan's views on economics in general and of his conception of the political role of the economist on the other. He criticizes welfare economics both for basing policy recommendations on theoretical concepts that have no observable empirical counterparts and for uncritical use of the concept of the social welfare function. Buchanan's ideal economist would offer reform proposals as hypotheses to be tested against the empirical observation of whether or not they would be accepted by either every participant in the economy or by an overwhelming majority of them. The other essays in this part develop these ideas in several directions. The economist's special expertise in studying individual motivation and behavior makes him most useful in the context of economic policy when he attempts to define good decision rules rather than prescribing "optimal" solutions to particular problems. This fits in with Buchanan's basic view of economics as being the study of exchange rather than the science of scarcity and choice. Choice is based on preferences and costs, both of which, according to Buchanan, are in principle unobservable. Exchange, on the other hand, is directly observable and can be studied by objective methods.
Part 2, "Individual Choices and Social Outcomes," contains a series of articles on externalities. One of the points that Buchanan stresses in these papers is that conventional welfare economics, at least by implication, takes an extremely naive view of the political process. Having diagnosed market failure in the presence of externalities, the welfare theorist would typically go on to show how the failure could be corrected by government action and then to recommend public interference with the market mechanism. But the recommendation is warranted only if the theorist also assumes that the government will in fact be acting according to the optimality principles that he has derived. There is no good reason to believe this, because the theorist has no hypothesis about the behavior of politicians and bureaucrats. Buchanan provides analytical examples to show that when such hypotheses are introduced, there is no a priori reason to expect government action to improve on the efficiency of the market allocation of resources.
The papers in Part 3, "Individual and Collective Choice," apart from an interpretive comment on Kenneth Arrow's Social Choice and individual Values (1951), contain a number of applications of the economic theory of individual choice to political behavior. One of Buchanan's basic methodological premises is that the study of man must be based on a consistent set of assumptions; man does not become a different species in the voting booth or in a political assembly from what he is in the market. One particularly interesting application of this idea is the theory that was developed in his famous 1965 article, "An Economic Theory of Clubs." The club is a voluntary association of individuals who join together to exploit the increasing returns aspect of public goods provision. The optimal club size balances the advantages of cost sharing against the disadvantages of congestion at the margin. This contribution was an important one both for bringing voluntary associations within reach of formal economic analysis and for demonstrating that although public goods provision does require some sort of collective action, the collectivity in question need not be located in the public sector. Other papers in this part are basically expositions of the individualistic approach to the analysis of political processes; Buchanan strongly emphasizes that politics is not a process of fact finding or identification of truth but rather of conflict resolution between individuals. This view has strong implications for the way in which economists conceive of their own role in society. Rather than attempting to find "optimal" solutions to economic problems they should concentrate on finding good decision rules, which all individuals and interest groups will find it in their own long-run interest to adopt for the solution of still unidentified conflicts over resource allocation.
The essays in Part 4 develop various aspects of the contractarian view of politics. In one essay Buchanan provides an interesting comparison between the social philosophies of Adam Smith and John Rawls (1971), arguing that Rawls is far from being the defender of the liberal welfare state that many economists have made him appear. According to Buchanan, the reason for this distorted view is that economists have concentrated their attention on the second of Rawls' two principles of justice, the maximin criterion for distributive justice, and neglected his first principle of equal liberties, which, according to Rawls himself, should be assigned lexical priority over the second. An important paper "Moral Community, Moral Order, or Moral Anarchy" essentially makes the point that the scope for government action depends on the degree of moral cohesion in society. Expanding the governmental sector may, by "relieving" private citizens of their social responsibilities, destroy the moral order.
The papers in Part 5, "Fiscal Economics as Political Economy," deal with public finance issues. These range from a very readable and interesting survey of the contributions of the nineteenth century Italian writers on public economics to an analysis of alternative political perceptions of the Laffer curve. A key role in the analysis in several of the papers is played by Buchanan's distinction between the choice of constitutional constraints and decisions made within a given set of constraints. In "Towards a Tax Constitution for Leviathan" (1977, written jointly with Geoffrey Brennan), the choice of a tax constitution is identified with the choice of basic principles of taxation, whereas the decision concerning the level of taxation will be made in the "postconstitutional stage." The authors then ask what kind of basic principles a representative taxpayer would like to see in the tax constitution, given that he believes that the state in the postconstitutional stage will behave like a Leviathan to maximize tax revenue within the limits set by the constitution. Their answer is that he will choose rules that in some sense minimize the maximal revenue that the government will collect in the postconstitutional state, and this will involve a narrow rather than a comprehensive tax base and a progressive rather than a proportional tax structure. Particularly the tax base result is in sharp contrast to those recommendations that are inspired by the optimal tax literature, where the idea is to minimize the deadweight loss of distortionary taxes for a given amount of tax revenue. Brennan and Buchanan argue that their model explains well the widely observed resistance of taxpayers to base broadening, even when this is coupled with proposals to reduce the tax rates.
Part 1 of Liberty consists partly of autobiographical material, partly of retrospective essays on the development of the public choice field. Part 2 "The Emergence of Order," (which includes two articles also published in Economics) raises, among other things, the question of whether "order" can be expected to emerge spontaneously out of the voluntary cooperation of individuals. As regards the market allocation of resources Buchanan answers the question in the affirmative; he also appears to deny that there can be a standard for judging market performance that exists independent of the market process itself, so that a system of unregulated markets is orderly or efficient by definition. As regards social institutions, on the other hand, there is no such tendency toward spontaneous order because of the public good nature of social institutions.
The essays in Part 3, "Explorations in the Theory of Justice," argue that justice should be interpreted basically as fairness, namely, as distributive rules that individuals themselves would consider to be just at the constitutional stage or behind a Rawlsian veil of uncertainty. This concept is fundamentally one of equality of opportunity, and Buchanan argues that the most important task for public distributional policy is to ensure that individuals are in equal positions with respect to their starting points. Within this framework he sees a case for taxes on intergenerational transfers and public financing of education as acceptable policy instruments. He also suggests that a progressive income tax might be part of the contractarian's just society. One essay discusses the ethical limits to taxation; here the focus is shifted away from the distribution of the tax burden to the justice of the burden in relation to what the individual gets back from the state in the form of, say, public goods. Here Buchanan suggests that individuals cannot morally be required to pay more in taxes than the amount that would make them prefer to secede from the community and form their own. In other words, every taxpayer should be left with a positive surplus in his transactions with the government.
Part 4, "The Political Economy of Debt and Deficits," gives Buchanan's views on the burden of the public debt and related matters. He argues that the prevalence of public deficits and increasing debt in modern Western societies reflects the breakdown of a moral rule against deficit financing, a breakdown that is due to John Maynard Keynes and his followers. Keynes launched his attack on the balanced budget rule with the best of intentions, wishing thereby to enable governments to stabilize the economy and to eliminate mass unemployment. However, where "Keynes totally failed was in his recognition that the long-standing rules for fiscal-monetary prudence were required to hold tribal instincts in check, and that once the Victorian precepts were eroded, the latter instincts would emerge with force sufficient to overwhelm all rationally derived argument" (Liberty, p. 192). The tribal instincts are the natural tendencies of politicians to disregard the interests of future generations and increase public spending by more than the increase in tax revenues. The public debt is a burden on future generations because, at least in the United States, the increased debt has gone to finance public consumption and not investment. But this state of affairs is inconsistent with long-run social and economic stability; "the welfare state guarantees its own demise upon the issue of its first dollar's worth of debt" (p. 210).
Part 5 contains a series of reflections on the relationship between the individual and the state. One of the central arguments here is that politics is concerned partly with law enforcement and collective action taken within the constraints given by the constitution, partly with changes in the constitution itself. For the latter type of decision the unanimity principle should be applied, but in reality this does not happen because modern political thinking does not recognize the fundamental distinction between the two types of public decisions. Buchanan is therefore strongly in favor of constitutional constraints on political choices. In the final essay "Political Economy and Social Philosophy" (1984), he identifies the three "demons" that plague the traditional economic approach to politics as the utilitarian calculus, the social engineering urge, and the elitist mentality that makes economists believe that they know something about the optimal social and economic order that the participants themselves do not.
Enough has been said about the contents of the two books to prove that they are rich in ideas and in stimulating analyses of important questions. They are also rich in strong and provocative statements, and not the least of their merits is that they force the reader to rethink his own position on fundamental issues. To achieve such a rethinking is obviously also one of the main concerns of the author, and it would be paying him a poor compliment simply to applaud him for his achievements. Accordingly, I shall discuss some of his ideas in more detail and not try to avoid criticism in the process.
III. The Theory of Individual Behavior
A central aspect of Buchanan's approach to economics is what he calls the individualistic postulate. This postulate says that the behavior of man must be explained by a consistent set of motivational assumptions, whether we want to analyze his behavior as a consumer, worker, voter, bureaucrat, or politician. One of the important contributions of the public choice school has been to emphasize that we do not have a complete positive theory of how the economy works until we have developed theories of man's behavior in a political and administrative context.(n4)
Buchanan has of course not been the only one to emphasize this point. Apart from the work of the public choice school itself, important early contributions were made by Anthony Downs (1957) and Joseph Schumpeter (1950). The postulate has a strong theoretical appeal. There would be something inherently unsatisfactory in switching between different sets of axioms for individual behavior in different arenas, given that it is the same individuals who perform in them. One should keep in mind, however, that the postulate is certainly not an innocuous one. One possibility is that people's behavior is to a large extent guided by norms, and that different norms apply in their roles as consumers, workers, voters, bureaucrats, and politicians. But the appeal of a unified theory is clearly so strong that it is at least worth giving it a try.
The economic behavior of the consumer (and worker) is based on the principle of utility maximization subject to a budget constraint. The arguments of the individual's utility function are assumed to be the quantities of the various goods consumed as well as the amount of leisure. The budget constraint requires the total outlay on consumer goods not to exceed the amount of labor income. There are well-known ways to extend the models to take account of time and uncertainty, thereby to explain saving and portfolio behavior, but these complications need not concern us here.
Can we use this model to explain the behavior of the individual in say, his role as a bureaucrat? We can clearly use it to explain occupational choice and therefore why some individuals prefer to become bureaucrats on the basis of expected lifetime income, job security, leisure characteristics, and so on. Can we also explain their behavior as bureaucrats? One possibility is to assume that individuals make bureaucratic decisions simply out of a concern for their own careers in terms of income, security, advancement, and so on. The behavior of bureaucrats, like the behavior of consumers, should on this assumption be understood in terms of the hypothesis of self-seeking behavior. In his foreword to Tullock's book, The Politics of Bureaucracy (1965 (1965)--reprinted as chapter 13 of Economics--Buchanan quotes Tullock's rephrasing of a famous statement by Adam Smith: "It is not from the benevolence of the bureaucrat that we expect our research grant or our welfare check, but out of his regard to his own, not the public interest" (Economics, p. 200). There are many instances in these books where Buchanan seems definitely to favor this rather narrow view of the motivation of politicians and bureaucrats, but there are also several examples of acceptance of a much broader view of human motivation. A general interpretation of his position would be that although a meaningful theory of political behavior can be developed on the basis of very general assumptions concerning human motivation, it is often a good research strategy to adopt a narrower view when analyzing particular problems. This should be a noncontroversial position. After all, it is the one we adopt all the time in conventional economic theory of market behavior. Although we admit in principle that the individual has one well defined preference ordering over private goods (including time-dated and contingent commodities) and public goods (including those that arise through external effects of private goods), we do not take all these complexities into account when studying, for example, labor supply behavior. If some noneconomist student of labor relations criticizes the theory for taking an unrealistically narrow view of the nature of man, we usually explain that the textbook model of labor supply is one chosen for analytical convenience, and that it does not in any meaningful sense represent economists' total vision of man as a worker. I imagine that Buchanan's different statements about political and bureaucratic motivation should be interpreted in a similar spirit. It cannot be an essential part of the positive theory of public choice to assume that behavior is always motivated by a narrow concept of self-interest, although such an assumption can sometimes be defended on the principle of Occam's razor. My own impression of some of the controversy that has arisen around this part of the contribution of the public choice school is that this methodological distinction has not been sufficiently clarified.
Buchanan does not himself, in these volumes, contribute very much to the positive theory of political behavior. In several contexts he makes assumptions about such behavior, especially in his discussions of the contractarian perspective to which I shall return below. But the assumptions are typically very general and defended mainly by reference to the work of other authors in the area, such as Tullock (1965) and William Niskanen (1971).(n5) There are no explicitly developed analytical models of the behavior of political man and only very general references to empirical evidence. In spite of his own strong advocacy of the positive approach, this has not been Buchanan's main field of interest within the area of public choice.
The obvious reason behind this choice of research strategy is that Buchanan has found other problems to be more interesting and challenging. But an additional reason may be that he has been uneasy about the conventional economic theory of individual choice. He repeatedly emphasizes the view that economics is the study of exchange and not the science of choice and optimization. One reason for this view is that technology and preferences are in principle unobservable and therefore not very attractive as building blocks for a scientific theory. Another is his conviction that only part of man's behavior comes within the legitimate domain of predictive economic theory. This is the part where individuals "are essentially passive responders to economic stimuli; they react; they do not choose" (Economics, p. 76). There is another part, however, that cannot be studied by means of operationally meaningful theory in the conventional sense. In this part the "objects for analysis are the choices of persons, which cannot be genuine choices and at the same time subject to prediction" (p. 76). I must confess that I find this point difficult to grasp. It is hard to see that there is a contradiction between assuming that behavior is predictable and acknowledging that the individuals involved make real choices. Voting behavior is a case in point. There can be little doubt that the act of voting in democratic elections is one where individuals have a strong feeling that they are making a genuine choice and not reacting to stimuli in a robot-like fashion; at the same time voting has proved to be one of the more predictable aspects of human behavior.(n6) The operational distinction between the two areas of human action is also troublesome; how do we know which type of decision belongs where? To say that this is an empirical question seems somehow beside the point, because we are then faced with a new problem: If our attempt at prediction does not work, is this because our predictive model is not good enough, or is it because prediction is impossible in principle? We seem to be dangerously close to a tautological explanation of the successes and failures of prediction.
This dual view is also somewhat puzzling as being apparently in conflict with the basic methodology of a unified theory of human motivation. If man--at least economic man--is envisaged as being essentially the same whatever the sphere he moves in, it is difficult to subdivide his decisions into those that can and those that cannot be predicted. One conventional way of achieving such a unified theory is to assume that all individuals have preferences defined over two vectors of variables, x and y. These preferences can be represented by the ordinal utility function u[v(x), y], where, for illustration, I have assumed weak separability between the two sets of variables. Let us think of x as being the set of ordinary consumption goods (including leisure) and y as various "social" variables beyond the individual's control as a consumer, such as the incomes of other families, levels of environmental pollution, and resources allocated to research in various universities. As a consumer the individual is constrained by the budget equation px = m, where m is nonlabor income. Maximizing u with respect to x subject to this constraint yields predictions about market behavior that are exactly the same as when v is maximized subject to the same constraint. Thus, market behavior can be interpreted as purely self-seeking, even though the consumer's complete preference ordering also gives weight to "social" variables. The point is that because the social variables do not enter his budget constraint as a consumer, there is no way that the consumer can reveal his complete preference ordering in the marketplace. Suppose now that this consumer is also a bureaucrat or a politician, concerned with the allocation of welfare checks or pollution permits or research grants. In this capacity he maximizes the function u with respect to y, and his budget constraint is now a different one. His checks, permits, or grants will probably have to satisfy some administratively determined budget constraint; in addition his public and private budget constraints may interact with each other through, for example, the effects of his bureaucratic achievements on his income m, or through the effects of his decisions on the market prices that he has to pay for his private consumption goods. Here there is a unified theory of human motivation in terms of one well-defined preference ordering; the broad applicability is achieved through the assumption that the individual faces different budget constraints in his different roles as a consumer and as a bureaucrat/politician. While the formulation clearly allows for self-seeking behavior in the context of public decision making, it also grants the possibility that the decision maker is motivated by concerns that go beyond his interest as a consumer.
I offer these remarks not as an interpretation of Buchanan's position, which, as I have shown above, is more complex, but as an alternative interpretation of the individualistic postulate. The challenge of the postulate could, according to this interpretation, be taken up by conventional theorists and incorporated in positive studies of public sector behavior without acceptance of Buchanan's views on the theory of choice, and this is in fact what has also happened in positive public choice theory. This does not imply, of course, that these views are without interest. And, as we shall see, they also influence his position on other issues, such as the role of applied welfare economics and the relationship between the individual and the state.
IV. Welfare and Efficiency
Does the market mechanism lead to an efficient allocation of resources? Most economists when confronted with this question, would probably answer it by reference to the theory of welfare economics. Welfare theory shows that under certain assumptions every competitive equilibrium is a Pareto optimum and that every Pareto optimum can be sustained as a competitive equilibrium. The standard exposition of these results begins by characterizing efficient allocations only by reference to the structure of the economy--the technologies of firms and the preferences and endowments of consumers--with no assumptions whatever being made about the organization of production and exchange. One may imagine an omniscient planner who tries to solve an optimization problem for society as a whole and then discovers that the solution to this problem can in fact be achieved by confronting consumers and firms with parametric prices that equate supplies and demands in all markets. The market economy can, according to some interpretations of this result, be seen as a "giant computer" or as a "solution algorithm" which, amazingly, manages to do just as well as the planner. Sometimes the market requires a little help from the planner in the form of selective taxes or subsidies or other forms of public sector interferences, and welfare economics also provides guidance with respect to the optimal forms of public policies to correct market failures. And by adding to the model a Bergson-Samuelson social welfare function the planner can design an optimal redistributional policy for the economy and thereby achieve not only efficiency but justice as well.
Buchanan objects to all this, and his objections are numerous. First of all, he objects to the use of a social welfare function. While acknowledging the importance of Arrow's impossibility theorem, his own objections to the welfare maximization approach is of a different kind. In Buchanan's view it is wholly inappropriate to attribute goals to society as such, and this is independent of the precise conditions required for the Arrow theorem to hold. This view is closely connected with his analysis of social decision processes, and I shall return to this below. At this point it is sufficient to note that he does not accept the derivation of distributional policies from the use of a social welfare function, whether this be the utilitarian, Rawlsian,(n7) or some other alternative specification.
On the other hand, Buchanan places great emphasis on efficiency as a standard by which to judge economic performance. Welfare theorists have traditionally been suspicious of efficiency as a goal when not related to the wider objective of welfare maximization. Bergson-Samuelson welfare maximization implies efficiency, but the satisfaction of the efficiency conditions does not imply that a maximum of welfare has been achieved. Does not the emphasis on efficiency then imply an inconsistency in Buchanan's position? Not necessarily. Buchanan's ideas about efficiency are not the same as the ones that are current in neoclassical welfare economics. He argues that because preferences and costs are unobservable, efficiency must be judged by the processes through which transactions are carried out, not by the results. A contract established by the voluntary cooperation between two or more parties (most of Buchanan's examples refer to two-party situations) is necessarily efficiency enhancing. The market allocation of resources is therefore efficient, provided that all contracts are voluntary and that all efficiency-enhancing contracts are in fact realized. As Buchanan himself acknowledges, this position may easily be interpreted as a pure tautology. He goes on to argue, however, that the set of contracts that will emerge depends on the institutional structure of society. Reforms designed to increase the degree of efficiency should therefore apply to the institutional structure, and with a change in this structure the set of contracts and the pattern of outcomes will change.(n8) In "Rights, Efficiency and Exchange: The Irrelevance of Transaction Costs," published in 1984 and reprinted as chapter 10 of Economics, he writes that this "suggests only that any allocation of resources that is to be classified as `efficient' depends necessarily on the institutional structure within which resource utilization-valuation decisions are made. This implication creates no difficulty for the subjectivist-contractarian who does not acknowledge the uniqueness of the resource allocation that is properly classified to be efficient" (p. 157).
To the extent that the "uniqueness" of the efficient allocation of resources is meant to refer to modern welfare economics, I find the latter part of this statement to be misleading. A Pareto optimum can be characterized as an allocation that maximizes a weighted sum of utility functions; consequently, the efficient allocation will depend on the weights chosen. The competitive equilibrium representation of this is that different weights correspond to different distributions of incomes or ownership rights. Institutional change can be interpreted as changes in the distribution of rights, so that in one sense the statement quoted is fully consistent with the position that follows very naturally from welfare economics. More generally, I feel that Buchanan tends to overstate the contrast between his own approach to problems of social efficiency and that of conventional neoclassical welfare economics. In another article (Liberty, p. 264) he argues that the claims to usefulness of Paretian welfare economics is built on the assumption that the economist knows the individual utility functions (and presumably firms' production functions as well). At least in one very important respect a more reasonable interpretation of the theory would be in a completely opposite direction: Under certain assumptions the decentralized market does as well as the omniscient planner, and because no omniscient planner exists, there is indeed a strong presumption that the market will lead to superior allocations in terms of efficiency. It is precisely because the enlightened planner realizes that he is very imperfectly informed about the economy that the main theorems of welfare economics are important for the design of policy. Thus, in this central area, welfare economists have been doing exactly what Buchanan has recommended economists to do, namely, identifying institutional structures that are conducive to efficiency. The proposal, strongly supported by many economists, to subject public projects to careful cost-benefit analyses before political decisions are made, could be interpreted in a similar vein. In its diagnoses of the possible sources of market failure I also feel that welfare economics has a clear advantage over Buchanan's rather vague position that what is, is efficient, relative to the structure of institutions.(n9)
Judging allocation mechanisms either by results or by the allocation processes themselves may not be a very fruitful contrast. Personally, I feel that to judge an economic system exclusively by its processes of resource allocation hardly makes sense; one must have some idea about the results to which the processes will lead. On the other hand, the alternative of evaluating efficiency (or social acceptability or justice) of an economic system solely in terms of outcomes also has its difficulties. One problem with this approach is that it is not always clear how one should draw the distinction between outcomes and processes, because the individuals in the economy may have preferences that are partly defined over the processes themselves. Thus, one may prefer competitive markets to central planning because markets result in more consumption goods for everybody, but one may also have a preference for having one's consumption goods delivered by the market rather than by the central planning bureau. Buchanan's emphasis that his interest lies in developing an economics for a free society may perhaps be seen as an expression of a similar point of view.
Perhaps the best-known part of Buchanan's criticism of welfare economics relates to the analysis of market failure. This does not stop with the diagnosis, it goes on to suggest remedies. In their analyses of the effects of externalities welfare economists may sometimes have fallen into the trap of arguing that if the market fails, the government is sure to do better. Their mistake has been to slip from the interpretation of their own theory as one of normative economics to interpreting it as positive theory, thereby making welfare economics and its applications much more progovernment and prointerference than could be justified on scientific grounds. Welfare economics is not a positive theory of government behavior, and to recommend public action to correct for market failure requires a positive theory of government behavior to convince us that there is in fact reason to expect that the government will do better than the market. Sometimes, no doubt, it will do worse. One may object that this is a very simple kind of insight that everybody understands. But the fact that we do so is in no small measure due to Buchanan's development of the idea.
I for my part do not feel that welfare economics has been destroyed by this kind of criticism. On the contrary I feel that the welfare analysis of efficiency, market failure, and the design of public policies to improve efficiency has strengthened its foothold after the public choice criticism. On the one hand I believe that the theory has withstood the attack on the notion of efficiency; the alternative of defining efficiency in terms of any result of voluntary transactions relative to the institutional structure seems to me a much less powerful and precise tool for analyzing market outcomes. On the other hand the Buchanan attack on the naive view of the welfare theorist as the adviser to a benevolent Bergson-Samuelson despot has led economists to take a much more realistic and modest view of their own roles as policy advisors, and this can only lead to a better understanding--and better practice--of normative economics.
V. The Role and Function of the Public Sector
Buchanan's contributions to public economics or the theory of public finance is in many ways a radical departure from the more conventional approach to this area. The standard approach is to think of the field as having a positive and a normative part. The positive theory is concerned with topics like the effects that exogenous changes in tax rates and other instruments of public policy have on saving, investment, and labor supply. The normative part is concerned with the derivation of rules for optimal public sector use of these instruments. Buchanan's main interests have been with neither of these parts. When he has been concerned with positive theory, it has been with the individual's choices as a participant in the political process, either as a voter or as a bureaucrat or politician. And although he has been severely critical of conventional welfare economics, he has been very much concerned with normative theory in the form of the constitutional or contractarian approach to the theory of the state; the latter has been his main interest in recent years.
To understand the main point about this approach it is useful to go back to Buchanan's critique of the naive version of welfare economics, in which it is implicitly assumed that public sector agents behave in accordance with the rules for socially optimal or efficient decisions. To support a belief in this view one would have to assume either that these agents faced an incentive system that actually motivated them to act in this way or that their behavior could somehow be perfectly monitored by the public. Neither of the assumptions is totally convincing as realistic descriptions of how the public sector functions. Furthermore, given the size and complexity of the public sector (even on a considerably smaller scale than the present one), it is not even conceivable that the public sector could ever be made to function in this way. For a given constitutional framework that defines the basic tasks of the public sector, how would we expect public sector agents to behave? Buchanan's general answer to this, as I have already pointed out, is that they will act in accordance with their self-interest. He further argues that self-interest will typically lead them to expand the public sector beyond the limits that would be drawn up by an enlightened polity behind a Rawlsian veil of ignorance. Those limits seem to be close to those of the nineteenth-century "night-watchman state"; Buchanan refers to the "necessary political role involved in enforcing individuals' rights and contracts and in producing those goods and services that are inherently public or collective in nature, including the legal system itself" (Liberty, p. 254). But in reality "there seems to arise a `natural' proclivity for individuals, groups, and the political entrepreneurs representing them, to extend the range and scope of collective-political action beyond any conceivable publicness boundaries, if publicness is defined in any economically meaningful sense.... The arms, agencies, and authority of the state will be utilized to secure, or in attempts to secure, differential gains for members of political coalitions, with little or no regard for normatively appropriate boundaries on governmental action" (pp. 254-55).
As a positive theory of public sector expansion this is obviously not intended to be anything but a sketch of an argument. But as I have already noted, Buchanan himself does not really contribute a more detailed theory to explain the expansion of the public sector in the present century, the only real exception to this being the hypothesis that Keynes broke down the Victorian moral precepts with respect to the public debt (Liberty, ch. 17). One would have liked to see a more rounded statement of Buchanan's views on this issue; after all, the boundaries of the night-watchman state may have expanded over time, and it is not clear that all of the growth of the public sector should be ascribed to the natural proclivities of public sector agents. That being said, however, it should be emphasized that the development of a positive theory of public sector growth is not Buchanan's main concern.(n10) Rather, the question that preoccupies him is the following: How would rational voters wish to delimit the powers of the state, given this insight into political behavior? Although it must still be counted as a weakness of the argument that the positive theory does not rest on a firmer foundation, it is perhaps not an essential weakness. It is here that the contractarian perspective become important.
Suppose that you are entering into a contract with a carpenter about repairs on your house. You may already have a theory of the behavior of carpenters as being on average highly skilled, punctual, and honest. Would you write the contract on the assumption that the theory applied also in this particular case? Probably not; the careful homeowner would like to have a contract that protected him against the possibility of the carpenter having the opposite characteristics. Buchanan argues that the same principle applies to the relationship between the individual and the public sector in a democracy. The individual does not have to believe that all public sector agents yield completely to their natural proclivities to expansion. But as a reflective voter and critical taxpayer he will find it rational to act on the assumption that they do.
This has a number of interesting implications. At a very basic level it provides a foundation for a theory of the state. If we imagine individuals being at the constitutional stage, that is, in a situation in which they are to design guidelines for the public sector, they would on the one hand allocate certain tasks to the public sector that the market cannot do on its own, for example, supplying public goods. On the other hand they would also constrain the public sector by rules designed to curb the expansionary tendencies of bureaucrats and politicians. These constraints could take the form of requirements of budget balance, restrictions on the set of taxes that can be used, and so on. While from a "social engineering" point of view such constraints could easily be interpreted as unnecessary and misguided restrictions on the use of policy instruments, they might in fact be interpreted as rational implications of prior choices at the constitutional level.
To my mind the most interesting application of this idea is to the theory of taxation. Mention has already been made of the important 1977 article with Brennan as coauthor (Economics, ch. 24). Apparently in contrast to the authors themselves I am more convinced by this application as a positive theory of taxpayer attitudes and behavior than as a normative theory. As a normative theory--that is, as a theory of how rational taxpayers would wish to restrict the government's behavior at the constitutional stage--it seems to suffer from a somewhat arbitrary set of assumptions. Why would taxpayers constrain the government's tendency to revenue maximization through the set of taxes and tax bases rather than by direct constraints on expenditure?(n11) But as a positive theory it seems to have considerable ability to predict taxpayer attitudes to tax reform. Some observations about my own country may serve to illustrate this. Parties on the left of the political center seem on the whole to be more interested in economists' advice on tax reform than parties on the right. This could be due simply to the stronger belief of the former in social engineering, but it could also be because a lowering of the marginal cost of public funds could justify a further expansion of public expenditure. Several Norwegian economists have advocated a change of tax policy that would lower marginal tax rates on income, which are currently very high, and raise taxes on property, including housing, which are very low; optimal taxation arguments involving demand and supply elasticities have been invoked to support this recommendation. However, such proposals appear to be very unpopular, even among those who in other contexts have argued that the high marginal rates of income tax are the main source of efficiency loss in the tax system.
There is perhaps a more general point here. The contractarian approach can be interpreted both as a positive and as a normative theory. The normative interpretation, which, as I understand it, is the one that Buchanan mostly has in mind, is in principle straightforward. From one perspective it can be seen as a variant of the modern theory of the principal-agent relationship. The people or the constitutional assembly are the principals, and politicians or bureaucrats are the agents. The contract between them should be set up in such a way that the agents are motivated to act in the interests of the principal. In particular applications of this theory, it is necessary to add to the general formulation a set of more particular assumptions on exactly how the interests of the two parties to the contract are supposed to diverge. It is at this point that a theory of the behavior of the agents in the public sector becomes especially important for convincing applications of the theory. However, there is also a positive interpretation of the theory, as the tax example above illustrates. If the contractarian organization of society is in the best interests of its citizens, one would expect there to be strong social forces acting to establish it. I read Buchanan as saying that although this is indeed the case, contractarian designs are vulnerable mechanisms that have to be protected from being undermined by shortsighted politicians and special interest groups. In this interpretation one needs a positive theory of behavior for both parties to the contract: for the voters to explain their demands for constitutional rules and constraints and for the politicians and bureaucrats to explain their behavior under a given set of constraints.
Buchanan sees his work in this area as the continuation of an old tradition in the public finance literature, stemming from the work of Knut Wicksell (1896)(n12) and the Italian public finance theorists of the nineteenth century. At the constitutional stage the tasks assigned to the state by voluntary agreement would have to be done behind the veil of uncertainty; one would, for example, have to set up procedures for conflict resolution in ignorance of which side of the conflict one would occupy in some postconstitutional situation. But in expectational terms at least there should be gains for all, and no individual should be required to pay more in taxes than the value to him of the goods and services that the public sector provides (e. g., Economics, pp. 307-08). At this stage, therefore, constitutional decisions should be made on the basis of unanimous support. Evidently, however, not all decisions made in the public sector can be made on the basis of unanimous approval by the electorate. Given that a large number of decisions have to be made by politicians and bureaucrats within a set of rules that leaves them considerable discretionary power, the possibility arises that these decisions may sometimes go against the true interests of the electorate. If such cases become frequent and serious, there is time for constitutional reform to limit the power of public sector agents. It is in this light that one must see Buchanan s support for Proposition 13 and related measures.
What becomes the scope for distributional policies in this perspective? Wicksell himself was a radical egalitarian and emphasized that the appeal of the voluntary exchange theory is conditional on the initial distribution of income being in some sense fair. In Buchanan's work this problem is not given much attention, although, as I have noted above, he does advocate a concept of justice as fairness, with the important condition for fairness being equality of starting points. It is hard to say, on the basis of his discussion, how far this concept would take us in terms of tax and transfer policies for redistribution or the distributional impact of public goods provision. I think many readers will come away with the impression that distributional policies are given rather low priority in Buchanan's system, although the basis for this impression may be as much an apparent lack of concern with the issues as any explicit statements.
Buchanan's criticism of the conventional theory of public finance is closely interwoven with his scepticism toward the growth of the public sector as epitomized by the welfare state. This is indeed natural. On the one hand economists may have taken their basic value premises from the dominant political ideology of the times; on the other hand some of that ideology has come from the contributions of economists. However, whereas much of the criticism of the welfare state from more conventional economists has focused on specific problems and reforms, Buchanan's objections refer mostly to general principles and seldom to concrete policy issues. Given the fundamental nature of the problems that concern him, this is understandable, but it sometimes also makes it hard to see what the implications of his position are with respect to the more practical problems facing Western governments. As he himself says of the Italian school of public finance, he is more concerned with system building than with problem solving. One partial exception to this is contained in a 1981 paper entitled "Dismantling the Welfare State" (Liberty, ch. 16). Here he is concerned not with imaginary constitutional decisions, but with a more practical problem of reform as it would face, for example, the government of Sweden (which he mentions as a particularly good example of the kind of society that he has in mind). He identifies the welfare state primarily with its elaborate system of redistribution through the social security system. What is wrong with this system is that it generates large efficiency losses via taxes and benefit payments, it lacks fiscal discipline and operates with enormous deficits, and it mixes up actuarially fair pensions with redistribution. Buchanan notes that if there is a gain from dismantling the system, it must be possible to do this with everybody coming out with a net gain, and he proposes that this be done by paying everyone the amount of his net claim on the system and thereafter establishing the pension system on a sound actuarial footing. The discussion is interesting, but it is disappointing in not taking account of the arguments that could be advanced in favor of a social security system in the first place; thus, there is no mention of market failure in insurance markets through moral hazard and adverse selection. It is also characterized by rather facile empirical assumptions. For example, after having argued that there will be "demonstrable increases in the productivity of the economy due to the reductions in excess burdens," he says that the "productivity gain will allow the economy to carry forward the annual interest charges on the accumulated welfare state overload with a gradually decreasing weight relative to annual gross product in the economy" (Liberty, p. 182). This is a fairly precise empirical statement, but it contains no references to data or econometric estimates of the supply and demand effects involved. Whatever the chances are for getting a unanimous vote in favor of this type of reform in Sweden, a lot of careful conventional analysis would have to be done in order to support the claims made for it.
VI. Further Perspectives
What, in perspective, are Buchanan's main contributions to economics? Questions like this are sometimes answered by listing the number of "firsts" that the person in question has achieved; this typically involves giving credit for being the first to formulate a new theoretical problem or provide a satisfactory solution to an old one. It may of course also involve priorities in empirical research by the use of new data or by the application of new methods of econometric analysis. Buchanan can obviously claim a fair number of "firsts" in this sense. The development of the theory of clubs and of the tax implications of the state as Leviathan are good examples. But this kind of assessment does not really do justice to Buchanan's contribution. His main achievements have been to introduce his fellow economists to new ways of thinking about economics, in particular about the public sector and the interaction between economics and politics. Although many will have reservations about the positions Buchanan has taken on central issues both in economic theory and in its applications to public policy, there can be no doubt that his views have been very influential, not only among those who consider themselves as belonging to the Buchanan "school," but also among economists in general.
As I have already mentioned, a rather striking feature of Buchanan's work is the almost total absence of empirical orientation. Not only has he not done any empirical work himself, but he has made hardly any reference to the empirical work of others. It is not clear how this should be interpreted. One simple explanation is that he has found other things to be more important. However, there are also other explanations. One is his subjectivist attitude to the measurement of preferences and costs. Another is that the specific issues with which he has been concerned are not such as to lead directly to empirical work; system builders are less empirically inclined than problem solvers. Therefore it is remarkable that the general area of public choice, as it is presented, say, in the journal Public Choice, has a rather strong empirical flavor. Of course, this may simply indicate that other influences, for example, from the work of George Stigler and his associates, have also been important. But in addition it may be that the indirect influences of Buchanan's work have inspired empirical work by leading researchers to explore data that pertain to the intersection between economic and political decision making and that they would not have seen this topic as interesting if it had not been for his writings.
Much of Buchanan's work has been concerned with criticism of the conventional way of doing economics. Because changing the priorities of the economics profession has been so much part of his objectives, a natural way to assess his contribution is to discuss his success in this respect. Has he succeeded in driving out the three demons that plague the profession?
The first of the demons, it will be recalled, is the utilitarian calculus. This should be understood as something more general than the utilitarian sum-of-utilities criterion; any attempt to aggregate utilities via a social welfare function is in Buchanan's view illegitimate as presupposing an organic conception of the state, a conception that is inconsistent with the basic value premises of a democratic society. I do not myself think that this particular demon has been very much affected by Buchanan's criticism. The large literature on social welfare functions and social choice has grown up more in response to the work of Arrow-(1951) and Amartya Sen (1970) than to that of Buchanan. And although it is widely recognized that a social welfare function cannot in general be derived from individual preferences by majority voting or any other aggregation procedure in such a way as to represent the preferences of the collectivity, this does not imply that such a function is not a useful tool for analyzing fundamental tradeoffs between efficiency and justice in the analysis of economic policy. That this has in fact been generally recognized is shown by the enormous growth during the last couple of decades in the application of welfare economics to areas like taxation and trade policies and regulation and financial markets. This particular demon seems to me to be alive and well, and I predict that he will stay with us for a long time to come.
The second demon, the urge toward social engineering, is at present in a more precarious position. Although Buchanan's picture of the naive welfare economist comes close to being a caricature, there can be no doubt that his message in this area has been widely received and understood. The economics profession as a whole is clearly much more sceptical today than it was 20 years ago about the possibility of improving the efficiency of resource allocation by interfering with the price mechanism or by designing optimal planning schemes for the public sector. On the other hand, many will remain unconvinced about the wisdom of giving up the social engineering role altogether. Personally, I believe that there is a lot to be said for Buchanan's recommendation that the economist see himself as someone who proposes policy changes in the form of hypotheses to be subjected to the empirical test of political approval, although I find it difficult to extend this to unanimous acceptance. In current debates about policy we are clearly not in a situation that resembles the constitutional stage, and the political process must inevitably face the problem of evaluating increases in the utility or standard of living of some individuals against declines on the part of other individuals.
The elitist mentality demon is perhaps the most complex of the three. I find it easy to agree with Buchanan when he speaks about the dangers of intellectual and academic arrogance. I believe also that there is a growing realization among economists that what may appear to be obstacles to efficient resource allocation may actually be the results of conscious attempts to create social institutions that are more in line with individual preferences or transactions costs, the nature of which is only imperfectly understood by economists. This realization naturally leads to more modest attitudes with respect to economists' pretensions to show the way toward the good society, so that this again is a point where Buchanan is in line with the current mood in the profession.
Scattered through these two volumes are a number of more personal pieces in which Buchanan speaks about his training, his early years as an economist, and his struggles to gain acceptance for the public choice perspective.(n13) These make lively and interesting reading. He tells the fascinating story of how he arrived as a graduate student at the University of Chicago as what he calls a libertarian socialist, that is, as one who is attracted to socialism because he values liberty above all else, While his values survived, his convictions changed abruptly after he had attended the lectures of Frank Knight for six weeks. He realized then that a market economy was the economic system that was best suited to realizing his basic values, "that there are narrow limits on the potential for exploitation of man by man, that markets tend to maximize freedom of persons from political control, that liberty, which has always been his basic value, is best preserved in a regime that allows markets a major role." With this realization he had "quite literally, seen the light" (Liberty' p. 5). In his Nobel lecture (reprinted as chapter 21 of Economics) he recounts how by coincidence in 1948 he came across Wicksell's Finanztheoretische Untersuchungen (1896) in the University of Chicago library, and how this discovery inspired him to develop the voluntary exchange approach to public finance. Several chapters look back on his experiences with creating an institutional setting for research and teaching in the public choice area. Although in his recounts of his experiences one can occasionally detect some bitterness at the lack of understanding in venous parts of the academic establishment, I imagine that when looking back on his efforts, he must feel that they have been successful. Naturally, being a controversialist, he has arroused controversy. That too might be counted as part of his success.
(*) I am grateful to Anthony B. Atkinson and two referees for helpful comments on an earlier draft.
(n1) J. M. Buchanan, Economics: Between Predictive Science and Moral Philosophy. Compiled by Robert D Tollison and Viktor D Vanberg. Texas A & M University Press, College Station, 1987, 413 pp. J.M. Buchanan, Liberty, Market and State: Political Economy in the 1980s. New York University Press, New York, 1986, 278 pp.
(n2) A companion volume to Economics, also edited by Tollison and Vanberg, is Buchanan (1989a). This came out too late to be taken account of in this review, except to note that it nicely complements the former volume. It is especially valuable in reprinting a number of articles that were originally published m less accessible sources.
(n3) For appreciations of Buchanan's work as a whole see the articles by Dennis Mueller (1984), Anthony Atkinson (1987), and Thomas Romer (1988). The area of public choice or the new political economy has been surveyed from somewhat different perspectives in Mueller (1979), Bruno Frey (1978), and Robert Inman (1987).
(n4) This was also the central message in the now classic work of Buchanan and Tullock (1962).
(n5) A useful reference on the current state of the theory of bureaucracy is Albert Breton et al. (1988).
(n6) Of course, individual prediction may be difficult even thoughprediction succeeds at the aggregate level, but most economic prediction is concerned with aggregates anyway. A constructive critique of prediction must therefore relate to its successes and failures at the aggregate level.
(n7) Rawls (1971) clearly sees his two principles as parts of one unified theory of justice. While Buchanan as shown above is a strong adherent of the first principle of equal liberties, he appears by his attitude to social welfare functions not to accept the second principle (the maximin principle of distributional justice). It would have been interesting to see a clarification of his position on this point.
(n8) I cannot see that this resolves the tautology problem, however. To decide whether a change in the institutional structure is likely to enhance efficiency in the operation of the market system, one must appeal to a standard that is independent of the market transactions as such.
(n9) It is true that many applications of welfare economics, in their analyses of policy, derive conclusions that depend on some nonobserved or even nonobservable magnitudes; maybe this is Buchanan's point when he refers to the theorist's assumed omniscience. Although such conclusions are not directly applicable, they may nevertheless be useful in directing attention to the need for collecting information that is not currently available.
(n10) The study of the growth of the public sector offers many interesting challenges, whether one's perspective is that of the public choice economist or a more conventional one. For an interesting attempt to combine theory and empirical data in this area see Thomas Borcherding (1985).
(n11) The aim of California's Proposition 13 was to introduce constraints on taxes and tax bases. An interesting question is why the attempt to introduce such constitutional constraints was not done much earlier. One possible but not totally convincing, answer is that the Victorian requirement of budget balance made measures like Proposition 13 superfluous.
(n12) Wicksell's work was continued by his countryman Erik Lindahl (1919). The Lindahl theory has been very influential on recent work in the area of public goods, beginning with its reformulation by Leif Johansen (1963). While the voluntary exchange approach is very clearly stated by Wicksell (1896), the specific contractarian perspective is Buchanan s own development of this set of ideas.
(n13) In addition, a very nice autobiographical essay entitled "Better than Plowing" has been published as chapter 7 of Buchanan (1989b).
REFERENCES
ARROW, KENNETH J. Social choice and individual values. New Haven and London: Yale U. Press, 1951.
ATKINSON, ANTHONY B. "James M. Buchanan s Contributions to Economics," Scand. J. Econ., 1987, 89(1) pp. 5-15.
BORCHERDING, THOMAS E. "The Causes of Government Expenditure Growth: A Survey of the U.S. Evidence," J. Public Econ., 1985, 28(3), pp. 359-82.
BRETON, ALBERT ET AL., eds. "Villa Colombella Papers on Bureaucracy," Europ. J. Polit. Econ. 1988, 4, Extra issue.
BUCHANAN, JAMES M. Explorations into constitutional economics. Eds.: ROBERT D. TOLLISON AND VIKTOR D. VANHERG. College Station: Texas A&M U. Press, 1989a.
------. Essays on the political economy. Honolulu: U. of Hawaii Press, 1989b.
BUCHANAN, JAMES M. AND TULLOCK, GORDON. The calculus of consent. Ann Arbor: U. of Michigan Press, 1962.
DOWNS, ANTHONY. An economic theory of democracy. NY: Harper and Row, 1957.
FREY, BRUNO S. Modern political economy. London: Martin Robertson, 1978.
INMAN, ROBERT P." Markets, Governments, and the `New' Political Economy," in Handbook of Public Economics. Vol. 2. Eds.: ALAN J. AUERBACH AND MARTIN FELDSTEIN. Amsterdam: North-Holland, 1987, pp. 647-777
JOHANSEN, LEIF. "Some Notes on the Lindahl Theory of Determination of Public Expenditure," Int. Econ. Rev., 1963, 4, pp. 346-58.
LINDAHL, ERIK. "Positive Losung", in Die Gerechtigkeit der Besteuerung. Lund, 1919. (Partial translation as "Just Taxation--A Positive Solution," in Classics in the theory of public finance. Eds.: RICH ARD A. MUSGRAVE AND ALAN T. PEACOCK. London: Macmillan, 1958, PP. 168-76.)
MUELLER, DENNIS C. Public choice. London: Cambridge U. Press, 1979.
------ "Mueller on Buchanan," in Contemporary economists in perspective, Eds.: HENRY w. SPIE GEL AND WARREN J. SAMUELS. Greenwich, CT: JAI Press Inc., 1984, PP. 557-78.
NISKANEN, WILLIAM. Bureaucracy and representative government. Chicago: Aldine-Atherton, 1971.
RAWLS, JOHN. A theory of justice. Cambridge: Harvard U. Press, 1971.
ROMER, THOMAS. "On James Buchanan s Contributions to Public Economics," J. Econ. Perspectives, Fall 1988, 2(4), pp. 165-79.
SCHUMPETER, JOSEPH! A. Capitalism, socialism and democracy. 3rd ed. NY: Harper and Row, 1950.
SEN, AMARTYA K. Collective choice and social welfare. San Francisco: Holden-Day, 1970.
TULLOCK, GORDON. The politics of bureaucracy. Washington, DC: Public Affairs Press, 1965.
WICKSELL, KNUT. Finanztheoretische Untersuchungen. Jena: Gustav Fischer, 1896. (Partial translation by J. M. Buchanan as "A New Principle of Just Taxation," in Classics in the theory of public finance. Eds: RICHARD A. MUSGRAVE AND ALAN T. PEACOCK. London: Macmillan, 1958, pp. 72-118.)
